Assessing the health burden of air pollution in African cities is challenging due to limited monitoring and significant data gaps. Cities like Kampala, Uganda, lack dense reference networks, making long-term health impact assessments difficult. In this study, we use fine particulate matter (PM2.5) measurements from a reference grade monitor, a network of low-cost air quality sensor nodes (LCAQSN) and satellite-derived PM2.5 to assess air quality trends in Kampala (2019-2023). To address the missing data problem in the PM2.5 time-series, we applied a two-stage machine learning (ML) imputation technique. The ML results were assessed using 5-fold cross-validation, having good predictions with an R2 = 0.78. This approach significantly improved data coverage, uncovering an additional 7-928 days of PM2.5 exceeding the 24 h World Health Organisation Interim Target 1 (WHO IT1) across traffic, sub-urban, residential and urban background sites. The comparison of the complete reference and LCAQSN PM2.5 time-series to satellite-derived PM2.5 showed good agreement with relative differences within 17%. Using an updated health-risk assessment model, we calculated an annual mortality burden between 660 and 1010 adult premature deaths. Reference-grade data had the highest PM2.5 mortality burden, which was 6% greater than satellite-derived data and 20% greater than LCAQSN data. Our results show that a two-stage ML imputation method can effectively fill long-term data gaps in ground-level measurements and reveal actual PM2.5 levels and related health impacts. In the absence of ground monitoring, satellite data provide a useful alternative for capturing city-wide PM2.5 trends and related health impacts.

Background Household air pollution (HAP), from the use of polluting fuels for cooking, heating, and lighting, poses significant health and environmental risks, particularly in low-resource settings. The Community Household Air Pollution Prevention Programme (CHAP-PP) integrated a ‘household air pollution, health, and prevention’ module into Kenyan national Community Health Worker training, involving household-based education, awareness-raising, and discussions around air pollution, health, and clean energy. Objectives This evaluation examined residents’ perspectives on the programme, considering impacts on energy use and HAP exposures in the context of wider experiences and providing recommendations for improvement. Methods This qualitative descriptive study used semi-structured interviews and one focus group discussion with purposively selected household representatives, analysed using reflexive thematic analysis. Results Residents welcomed the programme, reporting enhanced knowledge around HAP-related risks, harm mitigation practices (e.g. improved ventilation), and clean fuels. The latter enabled clean fuel adoption for some, but residual challenges remained. Using the COM-B (Capability-Opportunity-Motivation-Behaviour) model we considered how the programme influenced communities’ clean fuel uptake. Participants proposed solutions to financial barriers precluding clean fuel use including subsidies, value-added tax removal, and ‘pay-as-you-go’ schemes for liquefied petroleum gas, some of which have since been implemented. Conclusions The CHAP-PP programme was felt to be successful, increasing HAP awareness and supporting harm mitigation and transitions to cleaner cooking. Findings have informed module rollout nationally and contributed to Kenya’s HAP prevention strategy. Ongoing efforts aim to improve affordability and scale clean cooking solutions, with future evaluations planned to assess long-term impacts on energy use and health outcomes.\

In sub-Saharan Africa, approximately 85

Introduction School meals in low- and middle-income African settings are often prepared on polluting biomass stoves. Conversions to transitional clean fuels such as liquefied petroleum gas (LPG) could provide health and economic benefits, but little is known about end-user perspectives in school settings. This study explored the perspectives of cooks and school leaders on the cooking transition in Rwandan schools. Methods Semi-structured qualitative interviews (n = 34) were conducted across 2024 and 2025 with cooks and school leaders from 17 boarding schools that transitioned from firewood to LPG under a government-led programme. Interviews took place 20 to 41 months after transition. Data were analysed using thematic analysis. Results All schools reported using the LPG stoves, whilst also continuing to use firewood. Participants viewed LPG positively but indicated they would not have adopted it without the external programme. Reduced smoke was a central theme, with respondents linking LPG to cleaner kitchens and less smoke-related health concerns. LPG reduced the physical demands of cooking and saved time. Practical challenges included the limited cooking capacity of LPG equipment provided, difficulties preparing certain foods, and pot-handling, likely due to limited involvement of end users in the early stages of designing the intervention. Fluctuating gas refill prices and inconsistent supply remained challenging for continued use. Initial safety concerns were common, but training and routine use increased confidence. Conclusions This research reveals important lessons for optimising institutional clean cooking scale-up. Measures that ensure fuel affordability, reliable delivery, appropriately sized and high-quality equipment, and staff training are essential for exclusive and sustained clean fuel use. Rwanda's experience offers insights for other countries pursuing similar transitions.

Fuel poverty is shaped by interacting social, environmental and housing conditions, yet these links remain underexplored at city scale. The analysis is framed as an ecological, cross-sectional assessment of spatial associations rather than as a causal proof of a closed feedback mechanism. This study examines the relationship between fuel poverty, deprivation, particulate air pollution and housing typology across 54 wards in Liverpool, UK. Ward-level fuel poverty and Index of Multiple Deprivation (IMD) data were integrated with 2023-2024 annual mean particulate matter (PM2.5 and PM10) from 58 low-cost air-quality sensors and classified housing types. Regression models were used to compare individual, additive and interaction effects. Fuel poverty ranged from 12.4% to 25.29%, while PM2.5 and PM10 frequently exceeded World Health Organization guideline values. IMD was the strongest individual predictor of fuel poverty (R-2 = 0.281, p<0.001 ). The preferred additive model including IMD, PM2.5, PM10 and housing type explained 43.5% of the variance, with Victorian Terraces emerging as a significant risk factor. Although interaction models suggested pollution-deprivation coupling, model selection and uncertainty diagnostics favoured the simpler additive specification. The findings support targeted retrofit, fuel-poverty and emissions-control policies in deprived urban neighbourhoods where inefficient housing and environmental stressors compound energy insecurity and where local action can contribute to more equitable urban sustainability.

In 2022, the Russian-Ukrainian war onset increased global fuel and food prices. Additionally, a multi-season drought in Kenya reached peak severity in 2022. We aimed to quantify the impacts of these climatic and energy-related shocks on energy/food security in an informal settlement in Nairobi, Kenya. We leveraged real-time smart meter data and administered surveys from September–October 2022 to 701 pay-as-you-go liquefied petroleum gas (PAYG LPG) customers. Using mixed effects (repeated monthly measurements) log-linear regression, we compared PAYG LPG consumption/payment behaviors before (July–October 2021) and during a > 15

To reduce household air pollution exposure, consistent use of clean cooking fuels (e.g. liquefied petroleum gas (LPG)) is needed. While several studies have investigated determinants of LPG adoption cross-sectionally, few have explored factors associated with its use over time. Two surveys were administered to primary cooks in three peri-urban communities: Mbalmayo, Cameroon; Obuasi, Ghana and Eldoret, Kenya, with an approximate twomonth period in between surveys. Multivariable logistic regression models assessed predictors of continued primary LPG use. Within two months, 15% (n = 94) of 625 households cooking primarily with LPG ceased using it as their primary fuel. Additionally, 25% (n = 38) of 149 households using LPG as a secondary fuel stopped cooking with LPG altogether. In the multivariable model, primary cooks reporting constant availability of LPG at retailers had 1.34 times the odds (95% CI:1.07-1.70) of continued primary LPG use as those who found LPG refills to be unavailable once a month or more. Additionally, households with double (OR = 1.11, 95% CI: 1.03-1.20) or triple-burner LPG stoves (OR = 1.19, 95% CI: 1.10-1.30) had significantly higher odds of continuing to use LPG compared with those using a single-burner stove. Furthermore, households that used their stove seven days during the previous week had significantly higher odds (OR = 1.19, 95% CI: 1.10-1.30) of continued LPG use. Lastly, households in which the household head was unemployed had lower odds (OR = 0.94, 95% CI: 0.89-1.00) of continued primary LPG use. These findings underscore the importance of supply reliability and stove functionality as key determinants of continued clean fuel use, beyond initial adoption.

Background:Over 70% of Africans rely on polluting sources of energy for cooking. There is a paucity of epidemiological evidence on the burden of cooking fuel-related burns (CRBs) among women and children in low- and middle-income countries. Objectives:We estimated the prevalence of CRBs and association with main fuel choice among primary cooks and children 0-5 years of age in peri-urban areas in Kenya, Cameroon, and Ghana. Methods:We conducted a multisite cross-sectional survey in Mbalmayo, Cameroon; Obuasi, Ghana; and Eldoret, Kenya. Standardized questionnaires were administered between April 2019 and February 2020 to primary cooks. Questions included sociodemographic characteristics, primary fuel choice, and experience of burns within the previous 12 months. Overall and site-specific prevalence of CRBs were calculated, and their association with primary cooking fuel type was determined. Results:Overall, 128 out of 1,240 primary cooks [10.3%, 95% confidence interval (CI): 8.7, 12.2] reported at least one CRB during the previous 12 months. Most primary cooks had been burned multiple times ( median   number     of   burns = 3 , interquartile range: 2-5). CRB prevalence among primary cooks in Mbalmayo (23.3%, 95% CI: 19.4, 27.5) was significantly higher than in Obuasi (3.3%, 95% CI: 1.7, 5.8) and Eldoret (3.2%, 95% CI: 1.7, 5.3). Among children, the overall prevalence of CRBs was 5.1% (95% CI: 3.7, 6.9; n = 42 ) and was comparable across sites: Mbalmayo, 6.5% (95% CI: 4.0, 10.0); Eldoret, 4.7% (95% CI: 2.5, 7.9); and Obuasi, 3.9% (95% CI: 1.9, 7.1). Overall, there was no significant difference in CRB prevalence among liquefied petroleum gas primary users compared with exclusive biomass users considering primary cooks (11.8% vs. 9.2%, p = 0.17 ) and children (4.4% vs. 5.5%, p = 0.95 ). Older age [adjusted odds ratio ( aOR ) = 0.6 ; 95% CI: 0.3, 0.9; p = 0.03 ] and higher income ( aOR = 0.3 ; 95% CI: 0.2, 0.5; p < 0.01 ) significantly lowered odds of CRBs. Conclusions:CRB prevalence among primary cooks between communities was high but was not related to the main choice of fuel for cooking across the selected study sites. Older age and higher income significantly reduced the risk of CRBs among both primary cooks and their children. https://doi.org/10.1289/JHP1095.

While transitioning from polluting cooking fuels (e.g. wood, charcoal) to cleaner fuels, like liquefied petroleum gas (LPG), can lead to time savings, the amount of time saved is uncertain due to minimal stove use monitoring (SUM) data. Approximately three months (mean:82 days (SD:41)) of SUM data from Geocene temperature sensors was collected from 186 households in Mbalmayo, Cameroon; Obuasi, Ghana and Eldoret, Kenya. Households exclusively using LPG (mean:1 h 22 min/day) cooked for two hours/day less than those stacking LPG and polluting fuels (3 h 19 min/day), and almost three hours/day less than those exclusively using polluting fuels (4 h 10 min/day). Financially insecure households exclusively using polluting fuels cooked for ~ 45 min longer (4 h 29 min) than financially secure households (3 h 45 min). During a 24-hour household air pollution (HAP) monitoring period, average cooking time was 38 min longer (3 h 48 min vs. 3 h 10 min) and households cooked nearly once more per day (3.63 events) than during the remaining SUM period (2.72 events). Longer cooking times among financially insecure polluting fuel users suggests that LPG access may disproportionately benefit poorer households via greater time savings. Households may cook for longer-than-normal when monitored for HAP.

Air pollution poses a significant threat to global public health, with African megacities facing its severe consequences due to rapid urbanization, industrialization, and transportation challenges. In Africa, air pollution is responsible for 1.1 million deaths annually, with household air pollution accounting for two-third and ambient air pollution one-third of this burden. However, these percentages are likely to change in the near future due to the projected rapid urbanization and industrialization in the region. In the next 25 to 50 years African megacities are projected to grow rapidly and therefore experience a significant increase in air pollution-related health risks. Poor policy prioritization, limited monitoring infrastructure and conflicting interests and priorities further complicate the problem. In this paper, the key drivers of air pollution are discussed in African megacities, including urbanization, industrialization, transportation, and energy use. Further it is highlighted that there are significant challenges and barriers, as well as a pressing need for air quality monitoring, coordinated policies and effective air quality management to ensure sustainable development, mitigate the adverse health impacts of pollution and improve the quality of life across the continent.

In sub-Saharan Africa, approximately 85% of the population uses polluting cooking fuels (e.g. wood, charcoal). Incomplete combustion of these fuels generates household air pollution (HAP), containing fine particulate matter (PM2.5 ) and carbon monoxide (CO). Due to large spatial variability, increased quantification of HAP levels is needed to improve exposure assessment in sub-Saharan Africa. The CLEAN-Air(Africa) study included 24-h monitoring of PM2.5 and CO kitchen concentrations (npm2.5 = 248/nCO = 207) and female primary cook exposures (npm2.5 = 245/nCO = 222) in peri-urban households in Obuasi (Ghana), Mbalmayo (Cameroon) and Eldoret (Kenya). HAP measurements were combined with survey data on cooking patterns, socioeconomic characteristics and ambient exposure proxies (e.g. walking time to nearest road) in separate PM2.5 and CO mixed-effect log-linear regression models. Model coefficients were applied to a larger study population (n = 937) with only survey data to quantitatively scale up PM2.5 and CO exposures. The final models moderately explained variation in mean 24-h PM2.5 (R2 = 0.40) and CO (R2 = 0.26) kitchen concentration measurements, and PM2.5 (R2 = 0.27) and CO (R2 = 0.14) female cook exposures. Primary/secondary cooking fuel type was the only significant predictor in all four models. Other significant predictors of PM2.5 and CO kitchen concentrations were cooking location and household size; household financial security and rental status were only predictive of PM2.5 concentrations. Cooking location, household financial security and proxies of ambient air pollution exposure were significant predictors of PM2.5 cook exposures. Including objective cooking time measurements (from temperature sensors) from (n = 143) households substantially improved (by 52%) the explained variability of the CO kitchen concentration model, but not the PM2.5 model. Socioeconomic characteristics and markers of ambient air pollution exposure were strongly associated with mean PM2.5 measurements, while cooking environment variables were more predictive of mean CO levels.

With a wide range of stoves and appliances available in the ever-evolving Kenyan cooking market, it is important to understand which options are the most cost, time and energy efficient to use. This information can help households to make more informed decisions about their energy use and policy makers to better understand which solutions to promote. Despite its importance, the existing literature offers scant evidence to guide optimal stove and fuel choices. In this research, we utilised controlled cooking tests to investigate the fuel required to cook six regularly prepared dishes on 10 prevalent stove and fuel combinations (including liquified petroleum gas, ethanol, charcoal, kerosene and electric appliances). We also tested the efficiency improvements from pre-soaking beans and using stovetop pressure cookers. We collected primary fuel cost data from across Nairobi in June 2023 and collated historical fuel prices from secondary sources spanning 2019-2023. The prices of liquified petroleum gas, charcoal and kerosene varied considerably by variables such as brand and location, whereas ethanol and on-grid electricity were more stable. The electric pressure cooker was the most cost- and energy-efficient device. For liquified petroleum gas and charcoal, combining pre-soaking beans with a pressure cooker substantially reduced fuel consumption, but was still costlier than the electric pressure cooker. The longitudinal comparison highlighted the dynamic nature of fuel prices in Kenya and how a household's cost-optimal cooking stack can change at short notice. These findings demonstrate how comparative affordability varies both temporally and spatially and can be heavily affected by wider market and policy incentives.

School meals across Sub-Saharan Africa are typically prepared using biomass on inefficient stoves, resulting in high air pollution levels that might affect learners and staff. However, there is a paucity of air pollution health-related research in African schools. This study, conducted in seven schools in Rwanda and four schools in Kenya, assessed 1) levels of carbon monoxide (CO) and fine particulate matter (PM2.5) in school kitchens, classrooms (three, at different distances from the kitchen), playgrounds and personal among learners and catering staff; and 2) the prevalence of acute air pollution health-related symptoms and knowledge and perceptions of air pollution among learners and staff. For Rwanda and Kenya respectively, median 24-h PM2.5 levels were 263 and 1480 μg/m3 for kitchens and 63 and 68 μg/m3 for classrooms. In Rwanda, median personal PM2.5 exposure levels were 354 μg/m3 for cooks and 86 μg/m3 for leaners. In Kenya, median personal PM2.5 exposures were 1280 μg/m3 for cooks and 99 μg/m3 for leaners. Median CO levels in the kitchens were 1.8 and 23 and for cooks 3 and 14.8 mg/m3 for Rwanda and Kenya respectively. Surveys with learners (n = 526 and n = 302), catering staff (n = 45 and n = 28), and teachers (n = 21 and n = 12) for Rwanda and Kenya, respectively, demonstrated a high prevalence of self-reported air pollution-related headaches, eye irritation, and cough. The elevated air pollution levels and associated prevalence of health issues underscore the urgent need to accelerate transition to clean energy in African schools.

Fuel stacking perpetuates the negative impacts of polluting fuels and limits the potential of clean cooking transitions. The study aims to identify drivers of fuel stacking amongst customers of a pay-as-you-go (PAYG) LPG product provided by MGas in the greater Nairobi area as a basis for designing interventions that reduce fuel stacking. We developed a quantitative telephonic survey tool (n = 1323) to holistically investigate fuel stacking, which was validated by a smaller number of qualitative semi-structured interviews (n =18). Both the survey and interview designs were informed by Perros et al.'s 2022 taxonomy of fuel stacking drivers. Results showed that the main driver of fuel stacking was the incompatibility of PAYG LPG with specific cooking processes that were conducted regularly. This was most frequently due to the expense of heating large quantities of water and cooking long-boiling foods with PAYG LPG - tasks that participants reported are better performed by other stoves and fuels. Participants also faced technical and service-related issues with broken equipment, payment delays and incompatible personal cookware that sometimes rendered them unable to use PAYG LPG. We found weak correlation between self-reported stacking and actual PAYG LPG fuel use. These findings show that a single fuel or cooking technology is unlikely to efficiently and consistently meet all a household's cooking and water heating needs, and that fuel consumption is not solely driven by stacking practices. Clean energy providers should consider incorporating multiple modern energy cooking services comprising of fuel, stoves and compatible cooking utensils (e.g., pots and pans).

Background Exposure to household air pollution from polluting domestic fuel (solid fuel and kerosene) represents a substantial global public health burden and there is an urgent need for rapid transition to clean domestic fuels. Gas for cooking and heating might possibly affect child asthma, wheezing, and respiratory health. The aim of this review was to synthesise the evidence on the health effects of gaseous fuels to inform policies for scalable clean household energy. Methods In this systematic review and meta-analysis, we summarised the health effects from cooking or heating with gas compared with polluting fuels (eg, wood or charcoal) and clean energy (eg, electricity and solar energy). We searched PubMed, Scopus, Web of Science, MEDLINE, Cochrane Library (CENTRAL), Environment Complete, GreenFile, Google Scholar, Wanfang DATA, and CNKI for articles published between Dec 16, 2020, and Feb 6, 2021. Studies eligible for inclusion had to compare gas for cooking or heating with polluting fuels (eg, wood or charcoal) or clean energy (eg, electricity or solar energy) and present data for health outcomes in general populations. Studies that reported health outcomes that were exacerbations of existing underlying conditions were excluded. Several of our reviewers were involved in screening studies, data extraction, and quality assessment (including risk of bias) of included studies; 20% of studies were independently screened, extracted and quality assessed by another reviewer. Disagreements were reconciled through discussion with the wider review team. Included studies were appraised for quality using the Liverpool Quality Assessment Tools. Key health outcomes were grouped for meta-analysis and analysed using Cochrane's RevMan software. Primary outcomes were health effects (eg, acute lower respiratory infections) and secondary outcomes were health symptoms (eg, respiratory symptoms such as wheeze, cough, or breathlessness). This study is registered with PROSPERO, CRD42021227092. Findings 116 studies were included in the meta-analysis (two [2%] randomised controlled trials, 13 [11%] case-control studies, 23 [20%] cohort studies, and 78 [67%] cross-sectional studies), contributing 215 effect estimates for five grouped health outcomes. Compared with polluting fuels, use of gas significantly lowered the risk of pneumonia (OR 0·54, 95% CI 0·38–0·77; p=0·00080), wheeze (OR 0·42, 0·30–0·59; p<0·0001), cough (OR 0·44, 0·32–0·62; p<0·0001), breathlessness (OR 0·40, 0·21–0·76; p=0·0052), chronic obstructive pulmonary disease (OR 0·37, 0·23–0·60; p<0·0001), bronchitis (OR 0·60, 0·43–0·82; p=0·0015), pulmonary function deficit (OR 0·27, 0·17–0·44; p<0·0001), severe respiratory illness or death (OR 0·27, 0·11–0·63; p=0·0024), preterm birth (OR 0·66, 0·45–0·97; p=0·033), and low birth weight (OR 0·70, 0·53–0·93; p=0·015). Non-statistically significant effects were observed for asthma in children (OR 1·04, 0·70–1·55; p=0·84), asthma in adults (OR 0·65, 0·43–1·00; p=0·052), and small for gestational age (OR 1·04, 0·89–1·21; p=0·62). Compared with electricity, use of gas significantly increased risk of pneumonia (OR 1·26, 1·03–1·53; p=0·025) and chronic obstructive pulmonary disease (OR 1·15, 1·06–1·25; p=0·0011), although smaller non-significant effects were observed for higher-quality studies. In addition, a small increased risk of asthma in children was not significant (OR 1·09, 0·99–1·19; p=0·071) and no significant associations were found for adult asthma, wheeze, cough, and breathlessness (p>0·05). A significant decreased risk of bronchitis was observed (OR 0·87, 0·81–0·93; p<0·0001). Interpretation Switching from polluting fuels to gaseous household fuels could lower health risk and associated morbidity and mortality in resource-poor countries where reliance on polluting fuels is greatest. Although gas fuel use was associated with a slightly higher risk for some health outcomes compared with electricity, gas is an important transitional option for health in countries where access to reliable electricity supply for cooking or heating is not feasible in the near term. Funding WHO.

In sub-Saharan Africa, four out of five people use biomass fuels for household energy, with associated health and environmental problems. Interventions for clean cooking transitions tend to simplify agency and adoption motivations, with knowledge gaps in policy design. Drawing perspectives from own research from Tanzania, we focus on four key dimensions: household energy needs, climate, health, and the policy context. Six recommendations are highlighted for future research to inform evidence-based policy. First, gender is intrinsically associated with energy use patterns contradicting the common narrative that modern energy technologies can empower women. Second, fuel stacking is very common, and higher quality data is needed to better assess health and climate impacts from the energy use mix. Third, fossil LPG results in lower climate impacts than biomass energy, especially in contexts with high rates of deforestation. This challenges the dichotomy of renewable and non-renewable energy. Fourth, we query the polarity of clean vs. non-clean fuels, with charcoal found to be less polluting indoors once the stove has been lit outside. Fifth, energy policy effectiveness may be increased by combining Pay-As-You-Go fuel technologies with social services and policies beyond the energy sector. Sixth; poverty exacerbates the challenges of making decisions on essential household expenditure. Hence, policies should address poverty to ensure widespread adoption of clean fuels. The policy perspectives presented here are relevant for low and middle-income countries where the majority of the population relies on biomass fuels for their household energy.

BACKGROUND:Relatively clean cooking fuels such as liquefied petroleum gas (LPG) emit less fine particulate matter (PM2·5) and carbon monoxide (CO) than polluting fuels (eg, wood, charcoal). Yet, some clean cooking interventions have not achieved substantial exposure reductions. This study evaluates determinants of between-community variability in exposures to household air pollution (HAP) across sub-Saharan Africa. METHODS:In this measurement study, we recruited households cooking primarily with LPG or exclusively with wood or charcoal in peri-urban Cameroon, Ghana, and Kenya from previously surveyed households. In 2019-20, we conducted monitoring of 24 h PM2·5 and CO kitchen concentrations (n=256) and female cook (n=248) and child (n=124) exposures. PM2·5 measurements used gravimetric and light scattering methods. Stove use monitoring and surveys on cooking characteristics and ambient air pollution exposure (eg, walking time to main road) were also administered. FINDINGS:The mean PM2·5 kitchen concentration was five times higher among households cooking with charcoal than those using LPG in the Kenyan community (297 μg/m3, 95% CI 216-406, vs 61 μg/m3, 49-76), but only 4 μg/m3 higher in the Ghanaian community (56 μg/m3, 45-70, vs 52 μg/m3, 40-68). The mean CO kitchen concentration in charcoal-using households was double the WHO guideline (6·11 parts per million [ppm]) in the Kenyan community (15·81 ppm, 95% CI 8·71-28·72), but below the guideline in the Ghanaian setting (1·77 ppm, 1·04-2·99). In all communities, mean PM2·5 cook exposures only met the WHO interim-1 target (35 μg/m3) among LPG users staying indoors and living more than 10 min walk from a road. INTERPRETATION:Community-level variation in the relative difference in HAP exposures between LPG and polluting cooking fuel users in peri-urban sub-Saharan Africa might be attributed to differences in ambient air pollution levels. Thus, mitigation of indoor and outdoor PM2·5 sources will probably be critical for obtaining significant exposure reductions in rapidly urbanising settings of sub-Saharan Africa. FUNDING:UK National Institute for Health and Care Research.

There is a tendency to approach multiple fuel use mainly as a problem. This has limited how it is understood and addressed. This article uses a mixed-methods approach to advance understandings of multiple fuel use. We apply social practice theory to analyse multiple fuel use through a survey (n = 354) and 32 qualitative interviews in peri-urban Dar es Salaam, Tanzania. We find multiple fuel use with charcoal to be common, mostly in combination with LPG. Cooking solutions are combined in a complementary manner for their material qualities, but also by social factors. Multiple fuel use enables effective use of different energy types, saving money, coping with low and fluctuating incomes, and securing household energy supply. Flexibility also helps women navigate time constraints and multiple responsibilities, cook for larger families, maintain food culture, and make nutritious food affordable. Characteristics such as income level and variability, housing type, family size and children's age, are analysed as social differences influencing cooking patterns. We find multiple fuel use to be more than simply "stacking" of new fuels: it is an expression of competence and innovation. Efforts for change must build on holistic understandings of people's lived realities and should acknowledge their competent ways of manoeuvring difficult life conditions to a greater extent. Continued use of polluting cooking solutions through multiple fuel use is a barrier to clean energy transition, but solutions are more likely to succeed if recognising the multiple reasons and benefits from practicing multiple fuel use, rather than focusing on its negative impacts.